44 THROUGH CHITRAL AND MASTUJ [Chap. II Identifica- tion of Hsiian- tsang's Shang-mi. HsOan- tsang's de- scription of Shang-mi. different heights. All kinds of crops are grown, pulse and wheat being particularly plentiful ; of grapes there is abundance. This country yields laminary orpiment which is obtained by digging with the pick in the hill-side and breaking up the rocks where it is embedded. The spirits of the mountains are wicked and cruel, often causing great mischief. A man does not pass there without first offering a sacrifice; then he may go and return in safety. But if he does not address prayers to them storm and hail assail him. The climate is cold ; the manners of the people are lively and impetuous, their nature upright Their customs are not regulated by ceremony; they are narrow in their views and only moderately industrious. The writing is the same as that of the Tu-ho-lo (Tukhara) country ; but the spoken language differs. The people mostly wear woollen garments. The king is of the Sakya race. He respects and venerates the law of Buddha; his subjects follow his example and are all animated by sincere faith. There arc two convents containing a small number of monks/ Both Cunningham and Vivien de Saint-Martin had expressed the view that in this passage Hsiian-tsang was describing Chitral which, as seen both from the wording of the Hsi-yu-chi and the Life, he did not personally visit.17 Their suggestion received the support of Sir Henry Yule who pointed out that the yellow arsenic or orpiment mentioned by Hsuan-tsang; is still a characteristic product of Chitral.18 The fact that Shang-mi, with the alternative name of Chii-wci, is mentioned also in the brief notice of the Tcang Annals discussed above had not escaped the attention of Sir Henry Yule; but with only an inadequate abstract of the notice before him,19 it was impossible even for that great pioneer of the historical geography of Central Asia to perceive that the name Shang-mi had in reality a more restricted application. That the information heard and recorded by Hsuan-tsang must in the first place have related to the Upper Yarkhun Valley or Mastuj is evident from the fact that he mentions the route leading to it only after having traversed the whole of Wakhan from west to east, and immediately before beginning his description of the Pamirs (Po-mi-lo). Thus it is certain that the route leading south from the Sarhad across the Baroghil to the Yarkhun headwaters is intended. The description of Shang-mi as a cold region, the relatively small area indicated, and the mention of only two convents among a population wholly Buddhist are all indications which point to a mountain territory of limited extent and resources such as the present Mastuj.20 On the other hand it must be acknowledged that the reference to laminary orpiment, which is still obtained in Chitral much in the fashion described by the pilgrim, suggests the possibility that Hsiian-tsang was led by his informants to comprise under the same designation other portions of Kashkar-Bala or possibly the whole of Chitral. The vague and incorrect fashion in which the term Chitral is nowadays often extended by strangers to Mastuj, though the latter is an independent chiefship and always kept distinct in local knowledge, would supply an exact parallel. It must further be remembered that at the date of Hsiian-tsang's journey both Mastuj and Chitral may well have acknowledged for the 3T Cf. Cunningham in J.A-S.B., xiv. p. 433; V. de Saint-Martin, Mtmoire analyt., p, 426. 18 See his Notes on Tokharisian, f.jR.A.S., N. S., vi. p. 114. 19 Taken from Ritter, Asien, vii. p. 582. 20 The estimates of circumference so frequently given by Hsiian-tsang must necessarily be very vague. But in this case the corresponding measurements given for neighbouring mountain tracts may provide a gauge. Thus the small mountain territories of Shighnan (SkihJfi-m) and Kuran (Ku-lang-nd) at the headwaters of the Kokcha, are both given a circuit of over 2,000 li. Thus Ta-mo-hsi-t'ieh-ti or Wakhan, which consists of a valley scarcely much longer than the Yarkhun Valley and probably more limited in its arable land and population, is given a length of 1,500-1,600 li; cf. Watters, Yuan Chwang, ii. pp. 278 sqq. It may also be noted that Wakhan, though rightly described as less productive and with a far more severe climate, is stated to possess more than ten Buddhist monasteries, as against the two of Shang-mi. The number for Shang-mi might well surprise us in a population so devoted to Buddhism, if this district were taken to include the whole of the relatively fertile and populous Chitral.